Evaluation of the Hepatotoxic and Hepatoprotective Effect Of

Evaluation of the hepatotoxic and hepatoprotective effect of Rwandese herbal drugs on (guinea pigs barbiturate-induced sleeping time) and (rat precision cut liver slices, PCLS) models Marie-Jeanne Mukazayire, V. Véronique Allaeys, Pedro Buc Calderon, Caroline Stévigny, Marie-Josée Bigendako, Pierre Duez

To cite this version:

Marie-Jeanne Mukazayire, V. Véronique Allaeys, Pedro Buc Calderon, Caroline Stévigny, Marie-Josée Bigendako, et al.. Evaluation of the hepatotoxic and hepatoprotective effect of Rwandese herbal drugs on (guinea pigs barbiturate-induced sleeping time) and (rat precision cut liver slices, PCLS) models. Experimental and Toxicologic Pathology, Elsevier, 2010, 62 (3), pp.289. 10.1016/j.etp.2009.04.005. hal-00592285

HAL Id: hal-00592285 https://hal.archives-ouvertes.fr/hal-00592285 Submitted on 12 May 2011

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Author’s Accepted Manuscript

Evaluation of the hepatotoxic and hepatoprotective effect of Rwandese herbal drugs on in vivo (guinea pigs barbiturate-induced sleeping time) and in vitro (rat precision cut liver slices, PCLS) models

Marie-Jeanne Mukazayire, Véronique Allaeys, Pedro Buc Calderon, Caroline Stévigny, Marie-Josée Bigendako, Pierre Duez

PII: S0940-2993(09)00171-7 www.elsevier.de/etp DOI: doi:10.1016/j.etp.2009.04.005 Reference: ETP50354

To appear in: Experimental and Toxicologic Pathology

Received date: 14 July 2008 Revised date: 30 January 2009 Accepted date: 22 April 2009

Cite this article as: Marie-Jeanne Mukazayire, Véronique Allaeys, Pedro Buc Calderon, Caroline Stévigny, Marie-Josée Bigendako and Pierre Duez, Evaluation of the hepatotoxic and hepatoprotective effect of Rwandese herbal drugs on in vivo (guinea pigs barbiturate- induced sleeping time) and in vitro (rat precision cut liver slices, PCLS) models, Experi- mental and Toxicologic Pathology, doi:10.1016/j.etp.2009.04.005

This is a PDF ﬁle of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its ﬁnal citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Evaluation of the hepatotoxic and hepatoprotective effect

of Rwandese herbal drugs

on in vivo (guinea pigs barbiturate-induced sleeping time)

and in vitro (rat precision cut liver slices, PCLS) models

Marie-Jeanne Mukazayire a,b ( ), Véronique Allaeys c, Pedro Buc Calderon c,

Caroline Stévigny a, Marie-Josée Bigendako b, Pierre Duez a

a. Laboratoire de Pharmacognosie, de Bromatologie et de Nutrition Humaine, Institut de

Pharmacie, Université Libre de Bruxelles, CP 205/9, Bd du Triomphe, Bruxelles,

Belgique. b. Centre de Recherches en Phytomédicaments et Sciences de la Vie, Institut de Recherche

Scientifique et Technologique (I.R.S.T.), B.P. 227 Butare, Rwanda c. Unité de Pharmacocinétique, Métabolisme, Nutrition et Toxicologie, Faculté de

Médecine, Université Catholique de Louvain, Bruxelles, Belgique

Accepted manuscript

( ) Laboratoire de Pharmacognosie, de Bromatologie et de Nutrition humaine, Université Libre de Bruxelles (ULB), Campus de la Plaine – CP 205/9, Bd du Triomphe, B-1050 Bruxelles, Belgium. Phone, 32 - 2 - 650.5283; Fax, 32 - 2 - 650.5430; [email protected]

Abstract

Precision-cut liver slices (PCLS) preserve the tissular organization of the organ and represent an in vitro model closer to in vivo conditions than hepatocytes cultures. As this may be an interesting tool for the investigation of hepatotoxic and protective effects but also for bioguided fractionations schemes, the usefulness of PCLS was compared with an in vivo test of liver function. Crude extracts derived from five herbs used in Rwanda for hepatoprotective activity were tested on CCl4-treated guinea pigs by the method of barbiturate-induced sleep modification. Ocimum lamiifolium, Crassocephalum vitellinum, Guizotia scabra and

Vernonia lasiopus leaves extracts allowed animals to recover barbiturate sleep duration in proportions of 88, 78, 61 and 34 %, respectively and Microglossa pyrifolia was found inactive. Dried methanolic extracts of the 5 plants were then tested in vitro on rat PCLS for protection against acetaminophen - induced hepatotoxicity. In this model, Guizotia scabra,

Microglossa pyrifolia and Vernonia lasiopus were found hepatotoxic by themselves and unable to prevent acetaminophen toxicity. The most active extract, obtained from Ocimum lamiifolium, was subjected to bioassay-guided fractionation by chromatography on Si-C18 to yield two quite active fractions.

From a single animal, at least 50 PCLS explants can be prepared, which allows testing a large amounts of samples, strengthening ethnopharmacological data on hepatoprotective medicinal plants and investigating hepatotoxic effects.

Accepted manuscript

Key words: Rat precision-cut liver slices (PCLS); hepatotoxicity; hepatoprotection; Ocimum lamiifolium; Crassocephalum vitellinum; Guizotia scabra; Vernonia lasiopus;

Microglossa pyrifolia; Guinea pigs; Sleeping time

1. Introduction

The liver, major organ of metabolisation and excretion, is susceptible to a number of pathologies, classified as cirrhosis, acute chronic hepatitis and hepatitis; causes of hepatic trouble include parasitic and viral infections, autoimmune diseases and intoxication with various xenobiotics, such as chlorinated solvents, alcohol, drugs, herbal medicines, peroxidized fatty acids, fungal toxins, industrial pollutants and radioactive isotopes (Evans,

2002). The predominant pathologies in a given country depend on the lifestyle and economical conditions. In Rwanda, viral hepatitis and their complications, cirrhosis and hepatic carcinoma, represent 80 % of all liver pathologies, the 9th cause of morbidity

(Musemakweli, 1999), and a number of plants are traditionally used to treat liver diseases

(Van Puyvelde L., 1988). Excepting vaccines and interferon -2b, which concern only viral infections, modern medicine is quite limited in preventing or treating hepatic diseases; the only drugs available are cholagogues, choleretics, drugs for cholesterolic lithiasis, N-acetylcysteine and flavo-lignanes obtained from Silybum marianum. This limitation of therapeutic options gives considerable interest to the search for active compounds from plants traditionally used for these affections (Evans, 2002). In this context, an ethnopharmacological inquiry of a series of herbalists and/or traditional healers for plants used as remedies for liver diseases was undertaken by our team in Butare, Rwanda, in 2004, and confronted with published documents from Rwanda (Van Puyvelde L. et al., 1977; Van Puyvelde L., 1988;

Rwangabo, 1993) and other countries (Demissew and N., 1994; Neuwinger, 2000). This led to Accepted manuscript the selection of five herbs locally used for hepatoprotective activity (Table 1),

Crassocephalum vitellinum, Guizotia scabra, Microglossa pyrifolia, Ocimum lamiifolium and

Vernonia lasiopus. Such ethnopharmacological reports however do not give information enough regarding the clinical efficaciousness or safety of herbs and there are risks possibly associated with uncontrolled/unsubstantiated use of herbal remedies. The present study is

intended to explore whether these herbs could have protective effect on hepatocytes and to give an orientation to find hepatoprotective compounds that may be present in the extracts; this first step should eventually lead us to the molecules responsible for the effect to investigate mechanisms of action on the liver.

In order to investigate hepatotoxic and hepatoprotective compounds, i.e. agents capable to prevent and/or reverse the effects induced by an hepatotoxicant (CCl4, D-galactosamine, acetaminophen, peroxides,…), a series of models have been developed, either in vivo

(histology, measurement of serum hepatic enzymes, barbiturate-induced sleeping time, prothrombine time, bromosulphaleine clearance) or in vitro (continuous cell lines, primary cultures of hepatocytes). Precision-cut liver slices (PCLS) preserve the tissular organization of the organ and represent an in vitro model closer to in vivo conditions than hepatocytes cultures (Morales et al., 1998; Vanhulle et al., 2001; Vickers and Fisher, 2004). As PCLS may be an interesting tool for the investigation of hepatotoxic and protective effects but also for bioguided fractionation schemes, their usefulness has been further investigated on the five ethnopharmacologically selected Rwandan herbal drugs, comparing to an in vivo test of liver function, barbiturate-induced sleeping time.

2. Material and methods:

2.1. Plant material

The leaves of Crassocephalum vitellinum (Benth) S. Moore (Asteraceae), Guizotia scabra

(Vis) Chiov (Asteraceae), Microglossa pyrifolia (Lam.) Kuntze (Asteraceae), Ocimum Accepted manuscript lamiifolium Hochst ex. Benth (Lamiaceae) and Vernonia lasiopus O. Hoffm (Asteraceae) were collected in the prefecture of Butare (South-Western Rwanda) in July 2004, air-dried in the shade and mechanically powdered. The plants were authenticated by Dr M.J. Bigendako,

Center of Research in Phytomedicals and Life Sciences, Butare, Rwanda and voucher specimens were deposited in the BRLU herbarium, Belgium.

2.2. Chemicals

Williams’ medium E (WME) and foetal calf serum (FCS) were purchased from Gibco BRL

(Middlesex, UK), acetaminophen from Janssen Pharmaceutica (Beerse, Belgium), insulin

(Actrapid HM) from Novo Nordisk (Bagsvaerd, Denmark), gentamicine sulfate, carbon tetrachloride, pentobarbital, N-acetyl cysteine (NAC), ß-glucuronidase, benzoxazolinone chlorzoxazone and 6-OH- chlorzoxazone from Sigma-Aldrich (Bornem, Belgium).

2.3. Extraction of plant material

2.3.1. Extracts for in vitro tests and phytochemical screening Each plant (50 g) was macerated for 24 h and exhaustively percolated with solvents of increasing polarities, N-hexane, chloroform, methanol and water. The resulting extracts were evaporated to dryness (reduced pressure, 40°C), yielding four crude extracts (Table 2) which were subjected to phytochemical analysis, the methanolic extract being also used for in vitro tests.

500 mg of the most active methanolic extract (Ocimum lamiifolium) were submitted to column chromatography on silica gel-C18 eluting with a gradient of water - methanol (100% water, 90:10, 80:20, 60:40, 50:50 and 100% methanol); 60 fractions of 10 ml were collected and combined upon thin-layer chromatography (TLC) analysis to yield 9 fractions (A to I).

2.3.2. Extracts for in vivo tests

Crude extracts were prepared for the 5 plants by a healer working in the dispensary of the

Institute of Scientific and Technological Research of Butare, Rwanda, following his Accepted manuscript traditional protocol. Briefly, 350 g of each plant were infused in about 2000 ml of tap water, filtered on a cotton cloth and dried under reduced pressure at 40°C to yield the extracts.

The traditional use of the five plants relying on aqueous decoctions, only the polar extracts

(aqueous decoction and methanolic leaves extracts) were selected for pharmacological testing.

2.4. Animals

Male Wistar rats, weighing 250 - 300 g, were purchased from Iffa-credo (Les Oncins, France), housed in individual cages in a temperature and light controlled room (12 h dark/light cycles) receiving a standard diet AO3 (U.A.R., Villemoisson sur Orge, France) and water ad lib.

Male and female healthy guinea pigs, weighing 500 - 600 g and raised in the Institute for

Research, Science and Technology (IRST, Huye, Rwanda). Before the experiment, all guinea pigs were fasted for 16 hours and, throughout the duration of in vivo experiments (3 days), they were fed with a thorough mixture of 2 edible leaves, Lactuca capensis (Thunb.) (Asteraceae) and Brassica oleracea (var. capitata) (Brassicaceae) and tap water. These plants are known to be edible vegetables; our experience shows that guinea pigs can be grown with this regimen without any apparent problem. As control animals were fed the same mixture, diet parameters should be under control. All experiments were conducted in accordance with international standards of animal welfare as recommended by the European Union on Animal

Care (CCE Council 86/609).

2.5. Measurement of hepatoprotective activity

2.5.1. In vitro tests on acetaminophen-treated precision-cut liver slices (PCLS)

Surgical procedures were carried out on fed rats under pentobarbital (60 mg/kg) anaesthesia.

The liver was perfused in situ with ice-cold Krebs-Ringer solution before slicing. PCLS (250 -

300 from the whole liver without distinction of lobes in oxygenated ice-cold Krebs-Ringer buffer by using a Krumdieck tissue slicer according to Accepted manuscript procedures previously described (Evdokimova et al., 2001). They were stored 30 min at 4ºC in WME containing FCS (10% v/v), glutamine (2 mM), insulin (100 nM) and dexamethasone

(10 nM) and then transferred to 25-ml vials containing WME (2 slices/4 ml) supplemented with glutamine (2 mM), insulin (100 nM) and gentamicin sulfate (50 ). PCLS were incubated in a shaking water-bath (100 cycles/min) at 37ºC under a continuous flow of O2 -

CO2 (95 % : 5%) and randomized to avoid any variability between slices that may come from localization in the liver lobes or size. After 1 hour preincubation, allowing fresh slices to recover ATP content, the slices were transferred to other vials containing fresh medium and supplemented with different combinations of acetaminophen (hepatotoxic agent, 10 mM), N- acetyl cystein (hepatoprotective agent, 20 mM) and / or plant extracts (1 mg/ml) for 24 h.

Unsupplemented control slices were sampled after this 24 h incubation (Control). As the PCLS test is intended for bioguided fractionation (i.e. the use of a bioassay to guide the purification and isolation of active compounds), and given the relatively low amounts of material to be purified, a preliminary study led us to select a sensibly low dosage for tested extracts (1 to 5 mg/ml); this is the same order of magnitude as acetaminophen (1.5 mg/ml) and NAC (3.3 mg/ml).

Liver slices were taken, washed twice in saline and sonicated immediately in 1 ml of 2 % perchloric acid. The intracellular ATP content was measured on neutralized perchloric acid extracts using the ATP Bioluminescence Assay Kit CLS II from Boehringer-Mannheim

(Germany). Protein measurement was performed in duplicate on sonicated PCLS by the method of Lowry et al, 1951 using bovine serum albumin as standard (Lowry et al., 1951).

For each incubation condition, 2 slices were tested with 2 ATP measurements per slice; samples were compared by ANOVA and pairwise comparisons were performed by a t-test with Bonferroni correction. For histological examination, control and treated PCLS were immersed in 10 % formaldehyde, embedded in paraffin, sliced with a microtome, layered on Accepted manuscript microscope slides and stained with haematoxyline-eosine. CYP2E1 activity was measured by incubation of PCLS with chlorzoxazone and HPLC measurement of its hydroxylated metabolite, 6-OH-chlorzoxazone (Wauthier et al., 2004). Briefly, 4 slices per experimental condition in 2 ml medium were added with 20 l of 10 mg/ml chlorzoxazone in methanol and incubated (37°C, 90 min); 300 l of medium were added with 1.2 IU ß7

glucuronidase in 100 l acetate buffer (0.3 M, pH 5) and further incubated (37°C, 2h). Upon addition of 150 l ZnSO4 (15 % in water) and 100 l benzoxazolinone (0.1 mg/ml in methanol, internal standard) and centrifugation (13000 g, 1 min), the supernatant was filtered

(PVDF 4mm, 0.45m) and analyzed by HPLC in the following conditions : column C18

BDS 3 m (LC Packings); mobile phase, ammonium acetate (3.9 g/l, pH 4.25) - acetonitrile

(84:16) at 1.5 ml/min; UV detection at 287 nm.

2.5.2. In vivo tests on CCl4-treated male and female guinea pigs

Guinea pigs were divided into groups of 8 animals each, including a control group (no treatment prior to determination of sleeping-time). Animals were inspected every 12 h for humane endpoints, including behaviour (abnormal movements, immobility), clinical signs

(swelled abdomen, signs of severe distress, diarrhoea, convulsions) and eventual signs of impending death (convulsions, lateral position, recumbence, tremor); none of the animals had to be withdrawn from the study during the 3-days experiments. At day-1, the CCl4 group was given orally 1.5 mg/kg carbon tetrachloride as a fresh mixture with an equal volume of peanut oil. The duration of the sleep was determined two days after the administration of carbon tetrachloride (day-3). In this model, pure compounds have previously been tested from 200 to

800 mg/kg (Van Puyvelde L. et al., 1989); reasoning that the extracts are most probably mixtures of active and inactive compounds, with unknown bioavailability, we selected a i.p. dose of 1 g/kg.

At day-1, the treated groups received the preparation orally in doses of 1 g extract/kg one hour Accepted manuscript prior to carbon tetrachloride. The second and third day, the same dosage of the preparation was given and the duration of sleep determined on the third day, thirty minutes after the administration of the last preparation. Animals received pentobarbital intraperitoneally (i.p.) in doses of 50 mg/kg (the barbiturate was dissolved in 0.1 N NaOH). For each animal, sleep duration was defined as the time between loss and recovery of righting reflex.

2.6. Preliminary phytochemical analysis

Phytochemical screening of the plants was undertaken using described methods (Rwangabo,

1986; Wagner and Bladt, 1996). The screening covered (i) alkaloids (Mayer and

Dragendorff’s reagents; silica gel TLC in toluene-ethyl acetate-diethylamine (70:20:10) with spraying by Dragendorff and iodoplatinate reagents); (ii) terpenoids -sterols (Liebermann’s reagent; silica gel TLC in toluene-chloroform-ethanol (40:40:10) with spraying by anisaldehyde-sulphuric acid; (iii) flavonoids (Shinoda’s reagent); (iv) anthocyans (HCl 2N);

(v) saponins (honeycomb froth persisting for at least 30 min after vigorous shaking of the water extract; Liebermann-Buchard’s test); (vi) tannins (ferric chloride reagent and gelatinesalt block test); and (vii) anthraquinones (Borntrager’s reagent).

3. Results

Crassocephalum vitellinum, Ocimum lamiifolium, Guizotia scabra, Microglossa pyrifolia, and

Vernonia lasiopus were selected according to a ethnopharmacological inquiry and their use in liver diseases (Table 1). The leaves of the five plants were extracted with different solvents and their extraction yields are presented in Table 2. Each plant was screened for the main classes of phytochemicals (Table 3), revealing the presence of alkaloids, flavonoids, terpenoids - steroids, saponins and tannins as major chemical constituents.

Figures 1 and 2 compare the effects of the five different methanolic extracts with NAC as reference hepatoprotective agent in acetaminophen-challenged PCLS incubated in

Williams’medium either with or without paracetamol. On PCLS Guizotia scabra, Accepted manuscript Microglossa pyrifolia and Vernonia lasiopus were found hepatotoxic by themselves

(reduction in ATP levels, 68.7 %, 89.7 % and 88.3 %, respectively) and unable to prevent acetaminophen toxicity (Figure 1).

Crassocephalum vitellinum and Ocimum lamiifolium were not hepatotoxic and protected the liver slices against the induced hepatotoxicity (Figures 1 and 2). NAC proved extremely

efficient in preventing acetaminophen damage; indeed NAC forms a nucleophilic adduct with the electrophilic toxic quinoneimine metabolite (NAPQI). The Crassocephalum vitellinum and Ocimum lamiifolium extracts were comparatively less efficient than NAC, which may come from a different mechanism of action or a lower dose of active compounds in the tested extracts. Fractionation of the Ocimum lamiifolium by column chromatography on C18- modified silica gel and elution by a gradient water : methanol yielded 9 subfractions, Ocimum

A to I that present different TLC profiles. These were similarly tested on PCLS (data not shown) but only the first 2 fractions were found to prevent acetaminophen toxicity (Figure 2). The experiment was repeated on a series of animals for the extracts yielding a positive response (Table 4). Histological examination of slices (Figure 3) shows that acetaminophen induces altered and necrotic cells with widened sinusoidal spaces and chromatolytic nuclei.

NAC, Crassocephalum vitellinum and Ocimum lamiifolium extracts protect the tissue against acetaminophen toxicity, nuclei being apparent on a dense and coloured cytoplasm with normal sinusoidal spaces. Acetaminophen toxic metabolite, N-acetyl-p-benzoquinoneimine, is mainly produced by the cytochrome CYP2E1; the influence of active extracts on CYP2E1 was investigated by measuring the metabolization of a tracer compound, chlorzoxazone

(Table 5). Whereas O. lamiifolium fraction A had no effect by itself on control slices, the C. vitellinum methanolic extract slightly reduced CYP2E1 activity, which may partly account for its protective activity. The formation of 6-OH-chlorzoxazone was severely decreased by acetaminophen treatment (3 % vs control), which could partly be prevented by C. vitellinum Accepted manuscript methanolic extract (40 %) and by O. lamiifolium fraction A (51 %). These preliminary data however show no dose-effect relationship and need further investigation.

In in vivo studies, liver injury induced by CCl4, the best-characterized system of xenobiotic- induced hepatotoxicity, is a commonly used model for screening the anti-hepatotoxic / hepatoprotective activity of drugs (Brattin et al., 1985; Williams and Burk, 1990; Recknagel,

1991). As pentobarbital is metabolised exclusively in the liver, the sleeping time after a given dose is a measurement of hepatic metabolism. After poisoning with carbon tetrachloride, liver damage causes a sleeping time increase (Table 6); in the presence of an hepatoprotective agent, sleeping time is wholly or partly restored (Wagner and Wolff, 1977). The Microglossa pyrifolia aqueous extract does not protect the animals from CCl4 toxicity; protection afforded by Vernonia lasiopus was extremely modest (34.1 %). Guizotia scabra, Crassocephalum vitellinum and Ocimum lamiifolium allowed recovering sleeping time by 61.4 %, 77.9 % and

88.2 %, respectively. These data partly confirm those obtained on PCLS with methanolic extracts in which Guizotia scabra, Microglossa pyrifolia and Vernonia lasiopus were found hepatotoxic by themselves, whereas Crassocephalum vitellinum and Ocimum lamiifolium protected the liver slices against the acetaminophen hepatotoxicity..

4. Discussion

By contrast to many other pathologies, the symptomatology of a number of hepatic troubles

(icterus) is evident for traditional healers who can easily evaluate the response to treatments and thus probably select efficient herbal medicines. Many studies have already led to the characterization of more than 170 constituents isolated from 110 plants belonging to 55 families, including terpenoids, curcuminoids, lignoids and flavonoids and cyanogenetic glycosides (Luper, 1998; 1999; Evans, 2002; Seef et al., 2002; Thyagarajan et al., 2002) and a huge number of plants from India, China and Europe have been reported to treat liver diseases and boost liver functions (Curcuma longa, Picrorrhiza kurroa, Camellia sinensis, Glycyrrhiza Accepted manuscript glabra, Bupleurum falcatum, Silybum marianum, Taraxacum officinale,…) (Kiso et al., 1984;

Sugiyama et al., 1998; Ram, 2001; Evans, 2002; Thyagarajan et al., 2002), including some species related to those of the present study, Ocimum gratissimum, Ocimum sanctum and

Crassocephalum crepidioides (Chattopadhyay et al., 1992; Effraim et al., 2002; Aniya et al.,

2005). By contrast, only relatively scarce data are available on the African traditional

pharmacopoeia and the five plants we selected for this study, according to an ethnobotanical enquiry, have not been previously investigated for hepatoprotective activity.

Experimentally the concept of liver protection strongly depends on the chosen toxicant and laboratory animal species, which may lead to the selection of plant extracts and compounds active on a given model but without clinical relevance. To maximize the chances to select active extracts, two biological models, in vitro tests on acetaminophen-treated rat liver slices and in vivo tests on carbon tetrachloride-challenged guinea pigs, were selected for their different toxicity mechanistic. CCl4 is metabolised by CYP2E1 in an extremely reactive radical, CCl3 that peroxidizes lipids, inducing hepatic lesions (Plaa and Hewitt, 1998).

Acetaminophen (paracetamol), a frequently used analgesic and antipyretic drug, is primarily metabolized by sulfation and glucuronidation to unreactive metabolites, whereas a minor but highly reactive N-acetyl-p-benzoquinoneimine is normally detoxified by glutathione.

Following overdosage, the glutathione pool gets depleted and this metabolite accumulates to produce liver injury (Kupeli et al., 2006). In vitro studies on acetaminophen-treated PCLS show that N-acetylcysteine (NAC), a major antidote in clinical cases of acetaminophen intoxications, prevents depletion of ATP; NAC also offers an hepatoprotective effect against other toxins known to deplete cellular ATP levels (Popat et al., 2002).

As both animal species investigated (rat and guinea pig) have different sensitivity to toxic action of chemicals and significant differences in xenobiotics metabolism, this widens the research for possible hepatoprotective mechanisms and compounds; the ultimate challenges Accepted manuscript remain however (i) the translation from hepatoprotection to a potential beneficial effect on chronic diseases such as viral hepatitis or cirrhosis, which is still quite undocumented; and (ii) the validation of such data on humans. On our 2 models, the Ocimum lamiifolium and

Crassocephalum vitellinum polar extracts were found particularly active by contrast with

Microglossa pyrifolia (no activity on both models) and Vernonia lasiopus (modest activity

only on the in vivo model). A discrepancy was observed between models for the extracts of

Guizotia scabra; this may come from differences in toxicant activity mechanism, in the composition of tested extracts and/or doses.

The O. lamiifolium methanolic extract is already known for antipyretic activities in mice

(Makonnen et al., 2003) and for the presence of a major flavonol O-glycoside, the quercetin

3-O-xyloseyl galactoside, and large amounts of flavone 5-O-glycosides (Grayer et al., 2002).

Crassocephalum vitellinum and Guizotia scabra proved active on ileon and uterus of guinea pigs and on the arterial pressure of rabbits (Chagnon, 1984). Sesquiterpene lactones and other terpenoids were reported from these two herbs (Zdero et al., 1991; Zollo et al., 2000).

Vernonia lasiopus is known for antiulcerous and analgesic activities (Johri et al., 1995). Some glucosides with antibacterial activity and antimalaric terpenoids have been isolated from

Microglossa pyrifolia (Rucker et al., 1994; Köhler et al., 2002).

The hepatoprotective ability of O.lamiifolium and C.vitellinum polar extracts may be connected with their flavonoids, since these phytochemical constituents, detected in our active extracts, have been implicated as the hepatoprotective factors in Crassocephalum crepidioides

(Asteraceae) on CCl4-induced toxicity (Aniya et al., 2005) and in Equisetum arvense L.

(Equisetaceae) on tacrine-induced toxicity in Hep G2 cells (Hyuncheol, 2004).

Further work is in progress to determine if any of these compounds may be responsible for the hepatoprotective activities of Crassocephalum vitellinum and Ocimum lamiifolium.

To our best knowledge, this is the first time that the PCLS model is proposed and used for the Accepted manuscript bioguided fractionation of active extracts. PCLS indeed represent an attractive model in the study of natural products, a fair number of tests being carried on the explants of a single animal. The PCLS however present the drawback of not being perfused, the liver slices simply bathing in the nutrient medium. Compared to in vivo conditions, in which the liver is continually perfused by blood circulation, sensibly higher concentrations of test compounds

consequently have to be used. Compared to in vitro tests on monolayer cell cultures, higher concentrations of tested compounds are also required as cells inside the tissue need the diffusion of compounds through external cell layers (including external layers of dead cells induced by the slicing process). Despite this shortcoming, the PCLS model allows strengthening ethnopharmacological data on hepatoprotective medicinal plants and investigating hepatotoxic effects while sensibly reducing the number of experimental animals.

Our results support the Rwandese traditional use of four of the five plants as antihepatitis remedies; they suggest in vivo hepatoprotective effects of Ocimum lamiifolium, Crassocephalum vitellinum, Guizotia scabra and Vernonia lasiopus aqueous extracts. Given the direct toxicity of Guizotia scabra, Microglossa pyrifolia and Vernonia lasiopus methanolic extracts on the liver slices model, these 3 herbs should be further evaluated; their eventual use in traditional medicine should be quite cautious and accompanied by a monitoring of the hepatic function. It is noteworthy that the 2 herbs that proved active and non toxic on our models, C. vitellinum and O. lamiifolium leaves, are edible vegetables or food additives in Tanzania (Copeland, 2004) and Ethiopia (Demissew and N., 1994), respectively; this gives some more indications of a probably safe medicinal use . Further studies are worthy being carried out on the active compounds of Ocimum lamiifolium and Crassocephalum vitellinum in order to assess their structure and elucidate their mechanism of action.

Acknowledgements

The authors are extremely thankful to the Belgium Technical cooperation (BTC), the Belgian Accepted manuscript Fonds National de la Recherche Scientifique and the Institute of Research, Science and

Technology (IRST) of Rwanda for financial support of this work. The technical help of M.

Faes and O. Vaillant is gratefully acknowledged.

References

Aniya Y, Koyama T, Miyagi C, Miyahira M, Inomata C. Free radical scavenging and

hepatoprotective actions of the medicinal herb,Crassocephalum crepidioides from the

Okinawa islands. Biol Pharm Bull 2005; 28: 19-23.

Brattin WJ, Glende EA, Jr., Recknagel RO. Pathological mechanisms in carbon tetrachloride

hepatotoxicity. J Free Radic Biol Med 1985; 1: 27-38.

Chagnon M. General pharmacologic inventory of Rwandese medicinal plants. J

Ethnopharmacol 1984; 12: 239-251. Chattopadhyay R, Sarkar S, Ganguly S, Medda C, Basu T. Hepatoprotective activity of

Ocimum sanctum leaf extract against paracetamol induced hepatic damage in rats.

Indian J Pharmacol 1992; 24: 163-165.

Copeland SR, 2004. Paleoanthropological implications of vegetation and wild plant resources

in modern savanna landscapes, with applications to plio-pleistocene olduvai gorge,

Tanzania. Graduate School - New Brunswick Rutgers, The State University of New

Jersey, New Brunswick, p. 423.

Demissew S, N. A. Some useful indigenous labiates from Ethiopia. Lamiales Newsletter

(Kew) 1994; 3: 5-6.

Effraim K, Jacks T, Sodipo OA. Histopathological studies on the toxicity of Ocimum

gratissimum leave extract on some organs of rabbit. Afr J Biomed Res 2002; 6: 21-25.

Evans WC. An overview of drugs with antihepatotoxic and oral hypoglycaemic activities. In: Accepted manuscript Evans WC, editor. Trease and Evans Pharmacognosy, W.B. Saunders, Edinburgh; 2002.

pp. 414-420.

Evdokimova E, Taper H, Buc Calderon P. Role of ATP and glycogen reserves in both

paracetamol sulfation and glucuronidation by cultured precision-cut rat liver slices.

Toxicol in vitro 2001; 15: 683-690.

Grayer RJ, Veitch NC, Kite GC, Paton AJ, Garnock-Jones PJ. Scutellarein 4'-methyl ether

glycosides as taxonomic markers in Teucridium and Tripora (Lamiaceae, Ajugoideae).

Phytochemistry 2002; 60: 727-731.

Hyuncheol Oea. Hepatoprotective and free radical scavenging activities of phenolic petrosins

and flavonoids isolated from Equisetum arvense. J Ethnopharmacol 2004; 95: 421-424.

Johri RK, Singh C, Kaul BL. Vernonia lasiopus and Vernonia galamansis: a medicinal

perspective. Res Ind 1995; 40: 327-328.

Kiso Y, Tohkin M, Hikino H. Mechanism of antihepatotoxic activity of glycyrrhizin.Effect on free radical generation and lipid peroxidation. Planta Med 1984; 50: 298-302.

Köhler I, Jenett-Siems K, Kraft C, Siems K, Abbiw D, Bienzle U, Eich E. Herbal remedies

traditionally used against malaria in Ghana:Bioassay-guided fractionation of

Microglossa pyrifolia (asteraceae). J Biosciences 2002; 57: 1022-1027.

Kupeli E, Orhan DD, Yesilada E. Effect of Cistus laurifolius L. leaf extracts and flavonoids

on acetaminophen-induced hepatotoxicity in mice. J Ethnopharmacol 2006; 103: 455-

460.

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol

reagent. J Biol Chem 1951; 193: 265-275.

Luper S. A review of plants used in the treatment of liver disease: Part 1. Altern Med Rev

1998; 3: 410-421.

Luper S. A review of plants used in the treatment of liver disease: Part 2. Altern Med Rev Accepted manuscript 1999; 4: 178-189.

Makonnen E, Debella A, Zerihun L, Abebe D, Teka F. Antipyretic properties of the aqueous

and ethanol extracts of the leaves of Ocimum suave and Ocimum lamiifolium in mice. J

Ethnopharmacol 2003; 88: 85-91.

Morales H, Taper H, Buc Calderon P. Thermic transition and glycolytic capacity as critical

events in the survival of rat liver slices after overnight cold hypoxic preservation. J Appl

Toxicol 1998; 18: 103-109.

Musemakweli A. Séminaire sur les plantes medicinales et la pharmacopée traditionnelle.

Compte-Rendu des Travaux IRST, Butare; 1999. pp.

Neuwinger HD. African traditional medicine. A dictionay of plant use and applications.

Stuttgart: Medpharm Scientific Publishers; 2000.

Plaa GL, Hewitt WR. Toxicology of the liver. Washnigton DC: Taylor & Francis; 1998. Popat A, Shear NH, Malkiewicz I, Thomson S, Neuman MG. Mechanism of Impila

(Callilepis laureola)-induced cytotoxicity in Hep G2 cells. Clin Biochem 2002; 35: 57-

64.

Ram V. Herbal preparations as a source of hepatoprotective agents. Drug News Perspect

2001; 14: 353-363.

Recknagel RO. Free radical damage and lipid peroxidation. In: Meeks RG, editor.

Hepatotoxicology, CRC Press, Boca-Raton; 1991. pp. 401-436.

Rucker G, Kehrbaum SK, Sakulas H, Lawong B, Goeltenboth F. Medicinal plants from Papua

New Guinea. Part III. Acetylated aurone glucosides from Microglossa pyrifolia. Planta

Med 1994; 60: 288-289.

Rwangabo PC, 1986. Recherche des substances chimiques susceptibles de justifier l'activité

biologique de quelques plantes utilisées largement en medicine traditionnelle Accepted manuscript Rwandaise. Departement Farmaceutische Wetenschappen, Universiteit Antwerpen,

Antwerpen.

Rwangabo PC. La médicine traditionnelle au Rwanda. Paris: Karthala et ACCT; 1993.

Seef LB, Lindsay KL, Bacon B, Kresina TF, Hoofnagle JH. Complementary and alternative

medicine in chronic liver disease. Hepatology 2002; 34.

Sugiyama K, He P, Wada S. Green tea suppresses D-galactosamine- induced liver injury in

rats. Biosci Biotechnol Biochem 1998; 62: 609-611.

Thyagarajan S, Jayaram S, Gopalakrishnan V, Hari R, Jeyakumar P, Sripathi M. Herbal

medicines for liver diseases in India. J Gastroenterol Hepatol 2002; 17 Suppl 3: S370-

S376.

Van Puyvelde L. Contribution to the study of Rwandese medicinal plants, Butare. 1988.

Van Puyvelde L, Kayonga A, Brioen P, Costa J, Ndimubakunzi A, De Kimpe N, Schamp N.

The hepatoprotective principle of Hypoestes triflora leaves. J Ethnopharmacol 1989; 26: 121-127.

Van Puyvelde L, Ngaboyisonga M, Mukarugambwa PCS, Kayonga A, Barabwiliza. Enquêtes

ethnobotaniques sur la médicine traditionnelle rwandaise. Tome I. Préfecture de Kibuye.

Butare: Université Nationale du Rwanda; 1977.

Vanhulle VP, Martiat GA, Verbeeck RK, Horsmans Y, Calderon PB, Eeckhoudt SL, Taper

HS, Delzenne N. Cryopreservation of rat precision-cut liver slices by ultrarapid

freezing: influence on phase I and II metabolism and on cell viability upon incubation

for 24 hours. Life Sci 2001; 68: 2391-2403.

Vickers AE, Fisher RL. Organ slices for the evaluation of human drug toxicity. Chem Biol

Interact 2004; 150: 87-96.

Wagner H, Bladt S. Plant drug analysis. A Thin Layer Chromatography Atlas. Berlin:

Springer-Verlag; 1996. Accepted manuscript Wagner H, Wolff P. Proceeding in life sciences: New natural products and plant drugs with

pharmacological, biological or therapeutical activity. Berlin: Springer-Verlag; 1977.

Wauthier V, Verbeeck RK, Buc Calderon P. Age-related changes in the protein and mRNA

levels of CYP2E1 and CYP3A isoforms as well as in their hepatic activities in Wistar

rats. What role for oxidative stress? Arch Toxicol 2004; 78: 131-138.

Williams AT, Burk RF. Carbone tetrachloride hepatotoxicity: an example of free radical-

mediated injury. Semin Liver Dis 1990; 10: 279-284.

Zdero C, Bohlmann F, King RM, Robinson H. Sesquiterpene lactones and other constituents

from Siegesbeckia orientalis and Guizotia scabra. Phytochemistry 1991; 30: 1579-1584.

Zollo PHA, Kuiate JR, Menut C, Bessiere JM. Aromatic plants of tropical central

africa.XXXVI.Chemical composition of essential oils from seven cameroonian

Crassocephalum species. J Essent Oil Res 2000; 12: 533-536.

Accepted manuscript

Legends to the figures

Figure 1

Effect of N-acetylcysteine (20 mM) and the methanolic extracts of Ocimum lamiifolium,

Guizotia scabra, Microglossa pyrifolia and Vernonia lasiopus (1 mg/ml) on acetaminophen challenged (10 mM) and control precision-cut liver slices.

Mean + SD of duplicate ATP measurements on 2 PCLS obtained from the same rat;

*** : p < 0.001, NS : p > 0.05 (ANOVA with post-hoc pairwise comparisons; t-test with

Bonferroni correction).

Figure 2

Effect of N-acetylcysteine (20 mM), of the Ocimum lamiifolium and Crassocephalum vitellinum leaves methanolic extracts and of 2 fractions obtained from the O. lamiifolium extract (1 mg/ml) on acetaminophen-challenged (10 mM) and control precision-cut liver slices.

Mean + SD of duplicate ATP measurements on 2 PCLS obtained from the same rat;

*** : p < 0.001, NS : p > 0.05 (ANOVA with post-hoc pairwise comparisons; t-test with

Bonferroni correction).

Figure 3

Effect of N-acetylcysteine (20 mM), of the Ocimum lamiifolium and Crassocephalum vitellinum leaves methanolic extracts on acetaminophen-challenged (10 mM) PCLS.

(haematoxyline-eosine staining). Accepted manuscript

Figure

Figure 1 :

Accepted manuscript

Mukazayire - Figure 1 1

Figure

Figure 2 :

Accepted manuscript

Mukazayire - Figure 2

50 µm

50 µm methanolic extractmethanolic (2 mg/ml) Acetaminophen (10 mM)

C. vitellinum

Acetaminophen (10 mM) 50 µm methanolic extractmethanolic (2 mg/ml) Acetaminophen (10 mM) O. lamifolium + 50 µm 50 µm

Accepted manuscript Control (24 h) + NAC (20 mM) Acetaminophen (10 mM) Mukazayire - Figure 3

Figure 1

(a) Traditional uses Hepatitis, gonorrhea, gastric ulcer asthma, dropsy strong fever, dysmenorrhea,Hepatitis, woman sterility, facilitates the deliverance of placenta, constipation and kids' diseases Hepatitis, malaria, against helminthes Hepatitis, wounds and kid’s constipation Hepatitis, malaria, rhumatism, gastric ulcer, wounds disinfection, eyes treatment, head and abdomen pain Local name Umusura Isununu Ishikashike Ivumo Umuhe

Accepted manuscript

(Benth)S.Moore (Lam.) Kuntze O.Hoffm (Vis )Chiov : Traditional uses of the herbs investigated. Data combined from the present work and Van Puyvelde, 1988 Table 1 Scientific name (family; organ) Ocimum lamiifolium Hochst ex. Benth (Lamiaceae; leaves) Crassocephalum vitellinum (a) (Asteraceae; leaves) (Asteraceae; leaves) Guizotia scabra (Asteraceae; leaves) Vernonia lasiopus Microglossa pyrifolia (Asteraceae; leaves)

Table Table

Table 2 : Extraction yields (m/m) for the 5 plants selected (50 g leaves + 500 ml solvent; sequential 24 h macerations with increasing polarity solvents)

Solvent Yield (% m/m) Crassocephalum Guizotia Microglossa Ocimum Vernonia vitellinum scabra pyrifolia lamiifolium lasiopus n-Hexane 1.4 1.6 2.5 3.2 7.4 Chloroform 1.6 4.9 5 3.2 41.1 Methanol 9 13.9 18 6.4 8.4 Water 8.6 11.9 15.9 21.7 7.8

Accepted manuscript

+- ++ +- +- +-

Tannins Anthraquinones + + + + + (b) (2cm) (2cm) (1.3cm) (1.5cm) (1.5cm)

(1cm) (2cm) (2.2cm) (1.7cm) (1.5cm)

-+ -+ -+ ++ - +/- + (d) (d) (d) (d) (c) --+ Flavonoids Anthocyans Saponins --++ --++ ---+ (e) + (d) (d) (d) (d)

Accepted manuscript

CPHCMWPHCMW P P W P P (a) - -+++- -+ ++++++-+ +++++++++-+ ++++++-+ Alkaloids Terpenoids and steroids P +/-+/-++++-+ : Results of the phytochemical screening Shinoda's reagent : orange coloration (flavones) reagent : violet coloration (flavonones) In brackets, height of foam column Tested material (P = powder, H N-hexane extract, C chloroform M methanol W aqueous extract) Shinoda's reagent : red coloration (flavonols ) Table 3 (a) (b) (c) (d) (e) Plant Crassocephalum vittelinum Guizotia scabra Microglossa pyrifolia Ocimum lamiifolium Vernonia lasiopus

Table Table

Table 4 : Effect of N-acetylcysteine (20 mM) and the methanolic extracts of Ocimum lamiifolium and Crassocephalum vitellinum (1 mg/ml) on acetaminophen- challenged (10 mM) precision-cut liver slices from different animals

ATP (nmoles/mg proteins)

Number of Mean (b) SD RSD (%) animals (a)

Control 25 6.9 (***) 2.7 39 %

Acetaminophen 24 1.5 1.0 64 %

Acetaminophen + 12 6.4 (***) 2.5 39 %

NAC

Acetaminophen + 6 3.2 (*) 0.4 13 % Crassocephalum vitellinum

Acetaminophen + 10 4.5 (***) 1.4 31 % Ocimum lamiifolium

Acetaminophen + 4 8.2 (***) 5.1 63 % Ocimum lamiifolium fraction A

(a) 2 slices per animal

(b) Kruskal-Wallis post-hoc pairwise comparisons versus treatment Acetaminophen (Bonferroni correction; ***, p < 0.001; *, p < 0.05)

Accepted manuscript

Table

Table 5 : CYP2E1 activity of acetaminophen-challenged (10 mM) and control precision-cut liver slices upon treatment with N-acetylcysteine, Crassocephalum vitellinum (methanolic extract) and Ocimum lamiifolium (fraction A) (triplicate measurements on 4 PCLS obtained from the same rat; RSD = 6.2 %)

6-OH-chlorzoxazone Treatment μg/mg prot % vs control Control (24h incubation) 10.1 100 % C. vittelinum (2 mg/ml) 7.4 73 % C. vittelinum (5 mg/ml) 9.5 94 % O. lamiifolium fraction A (2 mg/ml) 10.1 100 % O. lamiifolium fraction A (5 mg/ml) 9.6 95 % Acetaminophen (10 mM) 0.3 3 % Acetaminophen + C. vittelinum (2 mg/ml) 3.4 34 % Acetaminophen + C. vittelinum (5 mg/ml) 4.1 40 % Acetaminophen + O. lamiifolium fraction A (2 mg/ml) 5.2 51 % Acetaminophen + O. lamiifolium fraction A (5 mg/ml) 5.2 51 %

Accepted manuscript

Table

Table 6 : Effect of the aqueous decoctions on the duration of pentobarbital-induced sleeping time (CCl4-intoxicated guinea pigs)

Plants Sleeping time (ST) (min) % recovery (b) (a) (mean ± SD; n = 8) Control pentobarbital (60 mg/kg) 144 ± 3 -

Control CCl4 (Hepatotoxicant, 1.5 ml/kg ) 300 ± 5 - Crassocephalum vitellinum 179 ± 6 *** 77.9 % Guizotia scabra 204 ± 7 *** 61.4 % Microglossa pyrifolia 297 ± 6 NS 1.7 % Ocimum lamiifolium 162 ± 5 *** 88.2 % Vernonia lasiopus 247 ± 5 *** 34.1 %

(a) Treatments were compared by ANOVA with post-hoc pairwise comparisons (t-test with Bonferroni correction; *** : p < 0.001, NS : p > 0.05). Hepatotoxi STcant Antihepato STtoxicant (b) % recovery x 100 Hepatotoxi STcant Pentobarbi STtal

Accepted manuscript